1 /* SPDX-License-Identifier: GPL-2.0+ */
2 /*
3  *  charset conversion utils
4  *
5  *  Copyright (c) 2017 Rob Clark
6  */
7 
8 #ifndef __CHARSET_H_
9 #define __CHARSET_H_
10 
11 #include <linux/kernel.h>
12 #include <linux/types.h>
13 
14 #define MAX_UTF8_PER_UTF16 3
15 
16 /**
17  * codepage_437 - Unicode to codepage 437 translation table
18  */
19 extern const u16 codepage_437[128];
20 
21 /**
22  * console_read_unicode() - read Unicode code point from console
23  *
24  * @code:	pointer to store Unicode code point
25  * Return:	0 = success
26  */
27 int console_read_unicode(s32 *code);
28 
29 /**
30  * utf8_get() - get next UTF-8 code point from buffer
31  *
32  * @src:		pointer to current byte, updated to point to next byte
33  * Return:		code point, or 0 for end of string, or -1 if no legal
34  *			code point is found. In case of an error src points to
35  *			the incorrect byte.
36  */
37 s32 utf8_get(const char **src);
38 
39 /**
40  * utf8_put() - write UTF-8 code point to buffer
41  *
42  * @code:		code point
43  * @dst:		pointer to destination buffer, updated to next position
44  * Return:		-1 if the input parameters are invalid
45  */
46 int utf8_put(s32 code, char **dst);
47 
48 /**
49  * utf8_utf16_strnlen() - length of a truncated utf-8 string after conversion
50  *			  to utf-16
51  *
52  * @src:		utf-8 string
53  * @count:		maximum number of code points to convert
54  * Return:		length in u16 after conversion to utf-16 without the
55  *			trailing \0. If an invalid UTF-8 sequence is hit one
56  *			u16 will be reserved for a replacement character.
57  */
58 size_t utf8_utf16_strnlen(const char *src, size_t count);
59 
60 /**
61  * utf8_utf16_strlen() - length of a utf-8 string after conversion to utf-16
62  *
63  * @a:			utf-8 string
64  * Return:		length in u16 after conversion to utf-16 without the
65  *			trailing \0. If an invalid UTF-8 sequence is hit one
66  *			u16 will be reserved for a replacement character.
67  */
68 #define utf8_utf16_strlen(a) utf8_utf16_strnlen((a), SIZE_MAX)
69 
70 /**
71  * utf8_utf16_strncpy() - copy utf-8 string to utf-16 string
72  *
73  * @dst:		destination buffer
74  * @src:		source buffer
75  * @count:		maximum number of code points to copy
76  * Return:		-1 if the input parameters are invalid
77  */
78 int utf8_utf16_strncpy(u16 **dst, const char *src, size_t count);
79 
80 /**
81  * utf8_utf16_strcpy() - copy utf-8 string to utf-16 string
82  *
83  * @d:			destination buffer
84  * @s:			source buffer
85  * Return:		-1 if the input parameters are invalid
86  */
87 #define utf8_utf16_strcpy(d, s) utf8_utf16_strncpy((d), (s), SIZE_MAX)
88 
89 /**
90  * utf16_get() - get next UTF-16 code point from buffer
91  *
92  * @src:		pointer to current word, updated to point to next word
93  * Return:		code point, or 0 for end of string, or -1 if no legal
94  *			code point is found. In case of an error src points to
95  *			the incorrect word.
96  */
97 s32 utf16_get(const u16 **src);
98 
99 /**
100  * utf16_put() - write UTF-16 code point to buffer
101  *
102  * @code:		code point
103  * @dst:		pointer to destination buffer, updated to next position
104  * Return:		-1 if the input parameters are invalid
105  */
106 int utf16_put(s32 code, u16 **dst);
107 
108 /**
109  * utf16_strnlen() - length of a truncated utf-16 string
110  *
111  * @src:		utf-16 string
112  * @count:		maximum number of code points to convert
113  * Return:		length in code points. If an invalid UTF-16 sequence is
114  *			hit one position will be reserved for a replacement
115  *			character.
116  */
117 size_t utf16_strnlen(const u16 *src, size_t count);
118 
119 /**
120  * utf16_utf8_strnlen() - length of a truncated utf-16 string after conversion
121  *			  to utf-8
122  *
123  * @src:		utf-16 string
124  * @count:		maximum number of code points to convert
125  * Return:		length in bytes after conversion to utf-8 without the
126  *			trailing \0. If an invalid UTF-16 sequence is hit one
127  *			byte will be reserved for a replacement character.
128  */
129 size_t utf16_utf8_strnlen(const u16 *src, size_t count);
130 
131 /**
132  * utf16_utf8_strlen() - length of a utf-16 string after conversion to utf-8
133  *
134  * @a:			utf-16 string
135  * Return:		length in bytes after conversion to utf-8 without the
136  *			trailing \0. If an invalid UTF-16 sequence is hit one
137  *			byte will be reserved for a replacement character.
138  */
139 #define utf16_utf8_strlen(a) utf16_utf8_strnlen((a), SIZE_MAX)
140 
141 /**
142  * utf16_utf8_strncpy() - copy utf-16 string to utf-8 string
143  *
144  * @dst:		destination buffer
145  * @src:		source buffer
146  * @count:		maximum number of code points to copy
147  * Return:		-1 if the input parameters are invalid
148  */
149 int utf16_utf8_strncpy(char **dst, const u16 *src, size_t count);
150 
151 /**
152  * utf16_utf8_strcpy() - copy utf-16 string to utf-8 string
153  *
154  * @d:			destination buffer
155  * @s:			source buffer
156  * Return:		-1 if the input parameters are invalid
157  */
158 #define utf16_utf8_strcpy(d, s) utf16_utf8_strncpy((d), (s), SIZE_MAX)
159 
160 /**
161  * utf_to_lower() - convert a Unicode letter to lower case
162  *
163  * @code:		letter to convert
164  * Return:		lower case letter or unchanged letter
165  */
166 s32 utf_to_lower(const s32 code);
167 
168 /**
169  * utf_to_upper() - convert a Unicode letter to upper case
170  *
171  * @code:		letter to convert
172  * Return:		upper case letter or unchanged letter
173  */
174 s32 utf_to_upper(const s32 code);
175 
176 /**
177  * u16_strncmp() - compare two u16 string
178  *
179  * @s1:		first string to compare
180  * @s2:		second string to compare
181  * @n:		maximum number of u16 to compare
182  * Return:	0  if the first n u16 are the same in s1 and s2
183  *		< 0 if the first different u16 in s1 is less than the
184  *		corresponding u16 in s2
185  *		> 0 if the first different u16 in s1 is greater than the
186  *		corresponding u16 in s2
187  */
188 int u16_strncmp(const u16 *s1, const u16 *s2, size_t n);
189 
190 /**
191  * u16_strcmp() - compare two u16 string
192  *
193  * @s1:		first string to compare
194  * @s2:		second string to compare
195  * Return:	0  if the first n u16 are the same in s1 and s2
196  *		< 0 if the first different u16 in s1 is less than the
197  *		corresponding u16 in s2
198  *		> 0 if the first different u16 in s1 is greater than the
199  *		corresponding u16 in s2
200  */
201 #define u16_strcmp(s1, s2)	u16_strncmp((s1), (s2), SIZE_MAX)
202 
203 /**
204  * u16_strlen - count non-zero words
205  *
206  * This function matches wsclen() if the -fshort-wchar compiler flag is set.
207  * In the EFI context we explicitly need a function handling u16 strings.
208  *
209  * @in:			null terminated u16 string
210  * Return:		number of non-zero words.
211  *			This is not the number of utf-16 letters!
212  */
213 size_t u16_strlen(const void *in);
214 
215 /**
216  * u16_strsize() - count size of u16 string in bytes including the null
217  *		   character
218  *
219  * Counts the number of bytes occupied by a u16 string
220  *
221  * @in:			null terminated u16 string
222  * Return:		bytes in a u16 string
223  */
224 size_t u16_strsize(const void *in);
225 
226 /**
227  * u16_strnlen() - count non-zero words
228  *
229  * This function matches wscnlen_s() if the -fshort-wchar compiler flag is set.
230  * In the EFI context we explicitly need a function handling u16 strings.
231  *
232  * @in:			null terminated u16 string
233  * @count:		maximum number of words to count
234  * Return:		number of non-zero words.
235  *			This is not the number of utf-16 letters!
236  */
237 size_t u16_strnlen(const u16 *in, size_t count);
238 
239 /**
240  * u16_strcpy() - copy u16 string
241  *
242  * Copy u16 string pointed to by src, including terminating null word, to
243  * the buffer pointed to by dest.
244  *
245  * @dest:		destination buffer
246  * @src:		source buffer (null terminated)
247  * Return:		'dest' address
248  */
249 u16 *u16_strcpy(u16 *dest, const u16 *src);
250 
251 /**
252  * u16_strdup() - duplicate u16 string
253  *
254  * Copy u16 string pointed to by src, including terminating null word, to a
255  * newly allocated buffer.
256  *
257  * @src:		source buffer (null terminated)
258  * Return:		allocated new buffer on success, NULL on failure
259  */
260 u16 *u16_strdup(const void *src);
261 
262 /**
263  * utf16_to_utf8() - Convert an utf16 string to utf8
264  *
265  * Converts 'size' characters of the utf16 string 'src' to utf8
266  * written to the 'dest' buffer.
267  *
268  * NOTE that a single utf16 character can generate up to 3 utf8
269  * characters.  See MAX_UTF8_PER_UTF16.
270  *
271  * @dest:	the destination buffer to write the utf8 characters
272  * @src:	the source utf16 string
273  * @size:	the number of utf16 characters to convert
274  * Return:	the pointer to the first unwritten byte in 'dest'
275  */
276 uint8_t *utf16_to_utf8(uint8_t *dest, const uint16_t *src, size_t size);
277 
278 /**
279  * utf_to_cp() - translate Unicode code point to 8bit codepage
280  *
281  * Codepoints that do not exist in the codepage are rendered as question mark.
282  *
283  * @c:		pointer to Unicode code point to be translated
284  * @codepage:	Unicode to codepage translation table
285  * Return:	0 on success, -ENOENT if codepoint cannot be translated
286  */
287 int utf_to_cp(s32 *c, const u16 *codepage);
288 
289 /**
290  * utf8_to_cp437_stream() - convert UTF-8 stream to codepage 437
291  *
292  * @c:		next UTF-8 character to convert
293  * @buffer:	buffer, at least 5 characters
294  * Return:	next codepage 437 character or 0
295  */
296 int utf8_to_cp437_stream(u8 c, char *buffer);
297 
298 /**
299  * utf8_to_utf32_stream() - convert UTF-8 stream to UTF-32
300  *
301  * @c:		next UTF-8 character to convert
302  * @buffer:	buffer, at least 5 characters
303  * Return:	next codepage 437 character or 0
304  */
305 int utf8_to_utf32_stream(u8 c, char *buffer);
306 
307 #endif /* __CHARSET_H_ */
308